Recall@K for each rank profile against a sampled validation set. Methodology lives in demos/visual-rag/eval/; rerun with uv run python evaluate.py --samples 200 --seed 42.
Latest run: — · — samples
| profile | R@1 | R@5 | R@10 | R@20 |
|---|---|---|---|---|
bm25 | — | — | — | — |
colpali | — | — | — | — |
hybrid | — | — | — | — |
The cells render as — until the first eval run is committed. The harness in the repo regenerates this table on demand against the live corpus — see the README in /eval/.